Research map / Point clouds / Point cloud classification
Fast and Scalable LiDAR Data Generation for Autonomous Driving Simulation without Raycasting
Irfan Nafiz Shahan, Al-Mubin Nabil, Arpan Kusari
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace raycasting with spherical projection plus gnomonic depth (GnoSphere360) | LiDAR simulation (HDL-64E, 288K rays) · CARLA Town03 | mean frame time (ms): 89.6170 -> 2.5420 | Improved |
| use GnoSphere360 instead of hardware-accelerated OptiX at baseline ray count | LiDAR simulation (HDL-64E, 288K rays) · CARLA Town03 | mean frame time (ms): 1.2270 -> 2.5420 | Hurt |
| use GnoSphere360 instead of OptiX at 4.6M rays | Hypothetical-256 LiDAR · Hypothetical-256 (4.6M rays) | mean frame time (ms): 13.54 -> 5.60 | Improved |
| use GnoSphere360 instead of OptiX at 6.3M rays | Hypothetical-350 LiDAR · Hypothetical-350 (6.3M rays) | mean frame time (ms): 62.98 -> 17.27 | Improved |
| use GnoSphere360 instead of OptiX for run-to-run consistency | HDL-64E baseline · CARLA Town03 | variance across runs (ms^2): 0.0110 -> 0.0003 | Improved |
| use GnoSphere360 instead of OptiX for variance across frames | HDL-64E baseline · CARLA Town03 | variance across frames (ms^2): 0.0162 -> 0.0701 | Hurt |
| use GnoSphere360 instead of OptiX for point fidelity | HDL-64E baseline · NN distance to OptiX points | points within 0.20 m (%): 99.967964 | Improved |
| increase maximum LiDAR range limit before triangle culling (ablation) | GnoSphere360 · CARLA Town03 | mean frame time | Hurt |
| use a denser scene (Town10HD, about 21M triangles) instead of sparser towns (ablation) | GnoSphere360 · CARLA towns 01, 03, 04, 10HD | mean frame time: about 21M | Hurt |
| use a denser LiDAR model (more rays, 256 channels) instead of HDL-64E (ablation) | GnoSphere360 · Town03 LiDAR model ablation | mean frame time (ms): 6 ms | Hurt |
Closest papers by meaning
- GS-LiDAR: Generating Realistic LiDAR Point Clouds with Panoramic Gaussian Splatting (ICLR 2025 · similarity 0.89)
- LiDAR-RT: Gaussian-based Ray Tracing for Dynamic LiDAR Re-simulation (CVPR 2025 · similarity 0.88)
- NeRF-LiDAR: Generating Realistic LiDAR Point Clouds with Neural Radiance Fields (AAAI 2024 · similarity 0.87)
- SplatAD: Real-Time Lidar and Camera Rendering with 3D Gaussian Splatting for Autonomous Driving (CVPR 2025 · similarity 0.87)
- LidaRF: Delving into Lidar for Neural Radiance Field on Street Scenes (CVPR 2024 · similarity 0.86)
- Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
